What it is
Pseudonymization replaces identifying values with consistent substitutes, preserving referential integrity across a dataset (the same person always gets the same fake name) without exposing real identities.
When to use it
Data and privacy engineers use this when preparing a dataset for testing, demos, or sharing with a third party, where relationships between records must survive anonymization.
Gotchas & limits
Pseudonymization is reversible if the seed and mapping logic are known; it reduces exposure but is not equivalent to true anonymization for regulatory purposes.